gecko

It sounds like a lizard, but in reality, GECKO helps you with your driving. It analyses driver patterns and computes comfort scores, driver rating and theft probability

GPS Based Encapsulation of Car Keyprint using OBD

This project has three parts :

• The torque app : Torque App Torque
• The GECKO app : GECKO App GECKO
• The flask server script : Script

What it does

It allows you to compute the following, from driver data :

Comfort Score : How the passengers feel while sitting in the car Driver Rating : How well you are driving, based on mileage and fault values Theft Probability : The probability that you're not the one driving the car (requires extensive data)

How to install

Steps :

1. Get an OBD Scanner : OBD Tool
2. Change the python script value of app.run(hostname,port) to your server's hostname and port. If running on local, it will be 127.0.0.1,5000
3. Run the python script. This requires you to install python-firebase and pandas via pip
4. Open the torque app and go to settings. Set "data upload to webserver" to "ON"
5. Set the link of the webserver to http://(your-hostname):(your-port)/test
6. Set the user email address to your first name. (Bug : @ in the email will not work)
7. Attach the OBD Tool to your car and drive!
8. Let the data be gathered
9. Open the GECKO app and type in the name in step 6.
10. Enjoy!

How it works

• Driver Identification : Random Forests
• Driver Rating : AdaBoost Algorithm
• Mileage : Kalman Filters
• Theft Probability : Random Forests
• Comfort Score : Linear Regression

Credits

Development

1. Pranav Patil
2. Niharika Sachdev
3. Shubham Kokate
4. Deptii Chaudhari

Ideation

1. Shivom Thakkar
2. Saloni Goel
3. Vishva Iyer
4. Sabarinath S
5. Ravleen Kaur